Homebrew PCBs

On Tuesday 01 February 2005 01:02 pm, Alan Rothenbush wrote:

> I have now added a new criteria to the parts decision.  Aside from
> function, power, price, availability and so on,
>
>   the thing cannot be invisible.
>
> If it's invisible, I have to continue looking.  That's my new rule.

Heh.  :-)

> Alas, even the days of visible parts are numbered.

I don't know about that.  I see people who are embracing surface mount parts 
and gearing up to deal with all of that,  and I'm just NOT gonna go there!  I 
have plenty of parts,  a pile of boards to scrap yet,  and there's lots more 
out there,  if I can get my hands on it before it hits the landfill.  I seem 
to be accumulating parts at a much faster rate than I'm using them up,  too,  
so I'm not real worried about running out any time soon.

Sure,  a lot of the industry is headed that way.  But ALL of it?  I have my 
doubts about that.  Sure,  a lot of parts are getting to where they're only 
available in surface-mount packages,  but a lot of fairly generic (what I 
call "jelly beans") parts are still around,  and I tend to design and build 
with those anyhow.

> I went to a seminar on PCB design done by IBM about 10 years ago.  At that 
> time, they passed around a part that looked a lot like a gull-wing RF 
> transistor.  Only the leads were obviously not any metal I'd ever seen ..
> they looked almost like mylar, giving off a weird moire effect when the 
> light struck them just so.
>
> Only under a microscope was it shown that the thing was an IC, with LOTS of
> THIN leads coming out each side.  The exact numbers escape me, but I think
> the leads were .001" wide on .001" centers or something ridiculous like
> that.
>
> Most of the crowd wanted to know how IBM produced a PCB to support a device
> with that fine a pitch; I wanted to meet the poor R&D guy who had to solder
> the first one to the prototype.

Anyway!

IBM does do a lot of very weird stuff like that over the years,  so it's 
really no surprise.  Whether such a thing will ever become in any way 
"mainstream" I don't know.

I scrapped some boards out years ago,  and couldn't figure out what some of 
the things on it were.  There were resistor packs,  pretty obviously,  but 
those had an assortment of different values in it.  I guess it paid them to 
make these so that they could save on assembly costs in using ordinary 
resistors?  Dunno,  but they were still through-hole parts.  Then there were 
these square parts with aluminum cases.  Twelve leads or so,  if I'm 
remembering right.  Only IBM part numbers on them,  which isn't gonna do me a 
whole lot of good (anybody know of a cross-refeerence for those? :-)

So I "took one apart".

Here the major portion of the inside of this thing was filled with what 
appeared to be white rubbery material (silicone?),  and on top,  at each 
corner,  were these *teeny little* gold wires,  going to what appeared to be 
a chip,  one attached to each corner.  Measuring the pins with that in mind 
showed these to be transistors!

What possible advantage they could have had to build these parts is beyond me.